Surgeon, Not Disease Severity, often Determines the Operation for Acute Complicated Diverticulitis Mohammad S Jafferji,
MD,
Neil Hyman,
MD, FACS
The “best” operation in the setting of acute complicated diverticulitis has been debated for decades. Multiple studies, including a recent prospective randomized trial, have reported improved outcomes with primary anastomosis. The aim of this study was to determine whether surgeon or patient-specific factors drives the choice of operative procedure. STUDY DESIGN: Consecutive adult patients with sigmoid diverticulitis, requiring emergent operative treatment for acute complicated diverticulitis, from 1997 to 2012 at an academic medical center, were identified from a prospectively maintained complications database. Patient characteristics, surgeon, choice of operation, and outcomes including postoperative complications and stoma reversal were noted. The use of primary anastomosis and associated outcomes between colorectal and noncolorectal surgeons were compared. RESULTS: There were 151 patients who underwent urgent resection during the study period, and 136 met inclusion criteria. Eighty-two resections (65.1%) were performed by noncolorectal surgeons and 44 by colorectal surgeons (34.9%). Noncolorectal surgeons performed more Hartmann procedures (68.3% vs 40.9%, p ¼ 0.01) despite similar demographics, American Society of Anesthesiologists (ASA) classification, and Hinchey stage. Length of stay, time to stoma reversal, ICU days, and postoperative complications were lower in the colorectal group (43.2% vs 16.7, p ¼ 0.02). CONCLUSIONS: Although patient-specific factors are important, surgeon is a potent predictor of operation performed in the setting of severe acute diverticulitis. A more aggressive approach to primary anastomosis may lower the complication rate after surgical treatment for severe acute diverticulitis. (J Am Coll Surg 2014;218:1156e1162. 2014 by the American College of Surgeons)
BACKGROUND:
More than 300,000 inpatient admissions for diverticulitis occur annually in the US.1 Although most cases of diverticulitis may be treated successfully without surgery, severe disease often requires emergent or urgent surgical management.2 For decades, the Hartmann procedure (HP) has been considered by many to represent the standard surgical approach for complicated diverticulitis.3-6 However, a growing number of studies have suggested that most acute cases may be safely treated by primary anastomosis (PA) with or without a diverting ileostomy.6-11 Despite the abundant literature attesting to the safety of primary anastomosis, HP continues to be the most commonly chosen alternative operation performed Disclosure Information: Nothing to disclose. Presented at the New England Surgical Society 94th Annual Meeting, Hartford, CT, September 2013. Received September 25, 2013; Revised November 6, 2013; Accepted December 9, 2013. From the Department of Surgery, University of Vermont College of Medicine, Burlington, VT. Correspondence address: Neil Hyman MD, FACS, Fletcher 465, University of Vermont College of Medicine, Burlington, VT 05401. email: Neil. [email protected]
ª 2014 by the American College of Surgeons Published by Elsevier Inc.
in the acute setting in the United States.1,12 The HP requires at least 2 major intra-abdominal operations, and many patients never have reversal of their stoma.13,14 Patients with complicated diverticulitis are typically seen in the emergency department and are evaluated by on-call surgeons, who determine the treatment of choice based on clinical and radiologic assessment, as well as their training and experience. Our aim was to assess whether surgeon or disease-specific characteristics primarily drive the type of operation performed in the setting of severe, acute complicated diverticulitis. To achieve this aim, we divided the surgeons who take night call for adult surgical emergencies into 2 categories: colorectal surgeons vs noncolorectal surgeons. We hypothesized that colon and rectal surgeons would be more likely to perform primary anastomosis. We also aimed to determine if any observed variation in surgical practice was associated with differences in key outcomes.
METHODS Study design and subject selection We conducted a retrospective chart review of all adult patients admitted through the Emergency Department
1156
ISSN 1072-7515/14/$36.00 http://dx.doi.org/10.1016/j.jamcollsurg.2013.12.063
Vol. 218, No. 6, June 2014
Jafferji and Hyman
Abbreviations and Acronyms
ASA HP LOS PA PA-DI
¼ ¼ ¼ ¼ ¼
American Society of Anesthesiologists Hartmann procedure length of stay primary anastomosis primary anastomosis with diverting ileostomy
of Fletcher Allen Healthcare, the teaching hospital of the University of Vermont College of Medicine, from January 1, 1997 through December 31, 2012, with a diagnosis of acute sigmoid diverticulitis (ICD-9 Diagnosis Code 562.11), who underwent an emergent or urgent surgical procedure during the index hospitalization. Patients were excluded from further analysis if they were under 18 years of age, had diverticulitis other than the sigmoid colon, had an alternative diagnosis made based on final histopathologic examination, a history of previous colon resection, or instances in which the admitting surgeon differed from the operating surgeon. Patients who did not have clear radiologic and/or intraoperative findings of perforated diverticulitis (defined as Hinchey stages 1 to 4) and those who underwent a radiologic drainage procedure before surgery or laparoscopic lavage were also excluded. Demographic information including age, sex, and body mass index (BMI) were noted. The white blood cell count and heart rate on admission to the emergency department were recorded. American Society of Anesthesiologists (ASA) classification was obtained from preoperative anesthesia records; Hinchey stage and any intraoperative complications were abstracted from the dictated operative report. The admitting/operating surgeon was recorded, and specific surgeon characteristics, including fellowship training and years in practice, were obtained from the institution’s surgical department database. Surgical procedures were classified as a Hartmann procedure (HP), resection with primary anastomosis (PA), or primary anastomosis with diverting ileostomy (PA-DI), based on review of the operative report. Postoperative complications were obtained from the Surgical Activity Tracking System (SATS), a prospectively maintained and validated complication database. Details of this tracking system have been published elsewhere.15 Any admission to the ICU, reoperation, or postoperative drainage procedure by interventional radiology was tracked in the patient’s medical record. The length of stay (LOS) was obtained from the patient discharge summary. Outpatient clinic records were routinely reviewed to capture complications occurring after discharge and to identify stoma reversal and days to reversal as applicable.
Surgery for Complicated Diverticulitis
1157
Statistical analysis Categorical variables such as ASA class, Hinchey class, or operation type were subjected to chi-square analysis and/ or Fisher’s exact test. Continuous data such as age, length of stay, length of operation, white blood cell count, and heart rate were analyzed with a Student’s t-test. Statistical computation was achieved using MS Excel 2007 and Stata/MP 12.0. The study was approved by the Institutional Review Board of the University of Vermont College of Medicine.
RESULTS A total of 151 patients were admitted through the emergency department with a diagnosis of severe acute complicated diverticulitis and underwent a surgical procedure at index hospitalization during the study period. Fifteen patients were excluded, usually based on disease location, subsequent histopathologic review, or the use of preoperative radiologic drainage; 10 patients were excluded because they were treated with laparoscopic washout without resection (Fig. 1). All procedures were performed by 1 of 14 board-certified surgeons who took general surgery night call during the study period; 3 were general surgeons, 4 were fellowship-trained colon and rectal surgeons, and 7 had fellowship training in other specialties (minimally invasive surgery, surgical oncology, or trauma). The surgeons were in practice a median of 11 years (range 1 to 27 years) at the time of colectomy. There were 88 operations (64.7%) performed by general surgeons or noncolorectal fellowship-trained surgeons; 48 operations (35.3%) were performed by a colorectal surgeon. Patient characteristics including Hinchey stage were similar between the 2 groups (Table 1). A total of 10 patients were treated with laparoscopic washout and were excluded from further analysis. Noncolorectal surgeons performed HP in 56 of 82 of the resected cases; colorectal surgeons performed HP in 18 of 44 instances (68.3% vs 40.9%, p ¼ 0.01). Patients in both groups who underwent a PA or PA-DI were significantly younger (p ¼ 0.04) and had a lower Hinchey class (p ¼ 0.01) and a lower ASA status (p ¼ 0.03) than those who underwent HP. The outcomes in the patients cared for by the noncolorectal and colorectal surgeons are displayed in Table 2 and by operation in Table 3. Of the Hinchey stage 4 patients, 16 of 18 noncolorectal patients underwent HP vs 3 of 8 in the colorectal group. In general, there was no difference in duration of surgery, overall LOS, need for reintervention, and the percentage of patients ultimately undergoing stoma reversal. Patients in the colorectal group were less likely to have an ICU admission, had a shorter LOS,
1158
Jafferji and Hyman
J Am Coll Surg
Surgery for Complicated Diverticulitis
Figure 1. Patient population.
and fewer days to stoma reversal. There were 13.7% of patients who required reintervention by interventional radiology or reoperation in the noncolorectal group vs 8.3% in the colorectal group (p ¼ 0.07). There were 2 intraoperative complications among the general group and none in the colorectal group. One was a bladder injury and the other was a ureteral injury. Postoperative complications occurred in 43.2% and 16.7% in the general and colorectal groups, respectively (p ¼ 0.02). There was 1 death in the noncolorectal group. The HP had approximately twice the complication rate of either PA group (Table 3).
Table 1.
DISCUSSION We found that colorectal surgeons were more likely to perform primary anastomosis in the setting of acute complicated diverticulitis than noncolorectal surgeons, despite operating on similar patients. This more aggressive approach was associated with fewer complications, shorter length of stay, elimination of ICU stays, and a shorter time to stoma reversal. However, it was not our primary aim or intention to compare the proficiency of colorectal surgeons to general or other fellowship specialty surgeons, nor to necessarily claim any sort of superiority in care based on training.
Study Patient Characteristics, n ¼ 136
Characteristics
Median age, y Female, % Median body mass index, kg/m2 Hinchey classification, median, % I II III IV
General and other surgeons
Colorectal surgeons
p Value
61 36.4 27.8 3.0 2 22 46 18
56 50.0 27.6 3.0 4 12 24 8
0.32 0.07 0.62 0.96
Vol. 218, No. 6, June 2014
Table 2.
Jafferji and Hyman
General and other surgeons
Colorectal surgeons
p Value
8.5 20.5 13.7 181 61.4 124
7 0 8.3 169 68.2 92
0.86 0.03 0.07 0.63 0.24 0.04
Median length of stay, d ICU admission, % Reintervention, % Median length of operations, min Stoma reversal, % Stoma reversal, d
It may be the case that other groups of surgeons produce equivalent or even superior outcomes in patients with complicated diverticulitis. Certainly, the small number of surgeons overall and in each subcategory would preclude any confidence in making such a sweeping assertion. Rather, we sought to use this comparison to help discern whether surgeon-specific or patient factors are the primary drivers of the operative approach in the setting of acute complicated diverticulitis. Undoubtedly, both the condition of the patient and the experience of the surgeon play some role in surgical decision making. Not surprisingly, patients who were older, had more advanced local sepsis, and a higher ASA classification were more likely to undergo an HP than their younger, more fit counterparts. This is the sort of observation that typically confounds the interpretation of the comparative studies of HP with PA.16,17 But our finding that more than twothirds of patients with complicated disease underwent colorectal anastomosis in the colorectal group vs well less than half of a well matched cohort, suggests that surgeon characteristics are a potent predictor of the operation performed. Although most patients with PA in this series had a protective loop ileostomy (PA-DI), the time to stoma closure was much shorter than for those who underwent
Variable
1159
Outcomes of Surgery, n ¼ 126
Characteristics
Table 3.
Surgery for Complicated Diverticulitis
HP. Hartmann takedown requires a second major laparotomy with a significant risk of inadvertent enterotomy and serious complications.18,19 In a relatively large series from our institution, the overall postoperative morbidity of Hartmann takedown was 29%, with an associated mortality of 2%.20 Further, approximately 90% of patients with PA-DI in previous comparative studies undergo ileostomy closure, and up to one-half of HP patients end up with a permanent stoma.9,18 In this series, performing primary anastomosis did result in somewhat longer operative times, especially when a loop ileostomy was added, but there was no difference in operative times between the colorectal and noncolorectal surgeons overall. The somewhat longer time to perform the anastomosis and construct the loop ileostomy must be weighed against the potential challenges of colostomy creation. Creating a left-sided colostomy in the acute setting, especially in obese patients, can be challenging and painstaking.21 Several patients in the HP group did require reoperation for colostomy complications, generally ischemia and/or retraction. Performing an anastomosis is often technically easier, and there is usually less trouble creating a loop ileostomy than an end descending colostomy.22
Comparison of Operative Groups by Surgeon Training, n ¼ 126
HP
Surgeon training Noncolorectal Colorectal PA PA-DI HP PA PA-DI
Median age, y 63.8 58 57.5 59.7 53.9 Median BMI, kg/m2 26.7 26.5 29.6 29.6 27.5 Median admission heart rate, beats/min 102 95 98 108 96 Median admission white blood cell count, 103/mL 14 9 11 13 10 Median Hinchey class 3.18 2.67 3 2.8 2.3 Median ASA class 3 3 2.8 3.1 2.7 Median length of operation, min 141 186 217 136 177 ICU admission, % 17.5 0 3 0 0 Mean length of stay, d 10.5 10.5 14 13.5 5.5 Complications, n (%) 19 (22) 9 (10.5) 10 (11.7) 4 (9) 3 (5.7) Median stoma, d 170 e 118 109 e
55.7 25.8 109 12 3.2 2.5 194 0 6.7 1 (2) 77
p Value (noncolorectal vs colorectal) HP PA PA-DI
0.29 0.68 0.69 0.87 0.23 0.88 0.12 0.03 0.01 0.06 0.03
0.45 0.78 0.76 0.67 0.44 0.35 0.48 0.79 0.04 0.08 e
0.23 0.39 0.31 0.59 0.76 0.40 0.26 0.22 0.01 0.06 0.008
ASA, American Society of Anesthesiologists; BMI, body mass index; HP, Hartmann procedure; PA, primary anastomosis; PA-DI, primary anastomosis with diverting ileostomy.
1160
Jafferji and Hyman
Surgery for Complicated Diverticulitis
There were more than twice as many postoperative complications in the noncolorectal group overall. This seems to be partially related to choice of operation because HP had the highest complication rate among the 3 operations, but also is related to surgeon-specific factors; there was a clear trend toward higher complications in the noncolorectal group for each individual procedure studied, as demonstrated in Table 3. We were surprised that no patient in the colorectal group required a stay in the ICU. The ICU stays were largely restricted to the noncolorectal patients who underwent an HP. The HP is often relegated to patients with the worst septic complications and/or the highest risk patients, so this finding would be expected.16 Certainly, patients with free diverticular perforation can be hemodynamically unstable, and we would not wish to imply that intensive care stays can be eliminated in patients with complicated diverticulitis. Rather, it appears that most patients do not present in this manner and can be safely managed without an ICU stay. The HP remains the most frequently chosen option in the United States for patients requiring acute surgery for diverticulitis. In a review of almost 100,000 patients from the National Inpatient Sample, more than 57% of patients were treated with a colostomy.12 Despite the abundant evidence that has accumulated over recent years attesting to the safety of primary anastomosis in the acute setting, there has not been a corresponding increase in the use of PA in actual practice.1,12 Multiple case series have reported superior outcomes for patients undergoing PA.6-11,17,23-27 However, as we observed in our series, these analyses are typically plagued by selection bias because healthier patients with less complicated disease are preferentially selected to have PA. A prospective randomized trial of 62 well matched patients comparing HP with PA-DI for the management of perforated left colonic diverticulitis demonstrated similar outcomes after the initial colectomy but revealed fewer serious complications in the PA-DI group (0% vs 20%), less cost, shorter operative times, and shorter length of stay associated with the secondary stoma closure procedure. Further, the stoma reversal rate was 90% in the PADI group vs 57% with HP.28 We did not break down the stoma reversal rate by individual operation in this series. The HP can be a lifesaving operation and likely has an ongoing role in managing selected patients with complicated diverticulitis. For example, patients who are hemodynamically unstable before surgery may present difficulties in the diagnosis of a postoperative anastomotic leak and may not be able to tolerate a second septic insult if a leak occurs. Further, there seems little reason to subject a patient with disabling fecal incontinence or those who are
J Am Coll Surg
institutionalized with serious comorbid conditions to the risk of anastomosis. Indeed, even the colorectal surgeons applied this operation in almost one-third of cases. However, as a rule, there may be little justification for HP in many patients with complicated diverticulitis. There are several limitations to this study. First this was a retrospective study at 1 institution involving a limited number of surgeons. A multicenter study would help to increase generalizability as well as increase the statistical power of the comparison. However, we also believe that using only our patients, whose outcomes are recorded by a rigorous, standardized prospective complication tracking system, also has some very real benefits regarding the completeness, accuracy, and uniformity of identifying adverse events.29-31 In addition, the decision to perform PA is not the only dilemma in the present day management of patients with severe acute diverticulitis. It remains unclear whether or not a loop ileostomy is really required after sigmoid resection with anastomosis,6,23 and we did not address the role of laparoscopic washout in patients requiring urgent surgery for perforated diverticulitis.32,33
CONCLUSIONS In conclusion, we believe that there is an opportunity for improvement in the care of patients requiring acute surgery for complicated diverticulitis by the more liberal use of PA, with or without diverting loop ileostomy. Although the HP likely has an ongoing role in the surgical armentarium and our data do not provide a basis for deciding when it is best applied, there seems little doubt that PA is underused. A more aggressive stance toward PA in the setting of acute complicated diverticulitis may lower complication rates, decrease length of stay, and improve patient satisfaction by reducing the incidence and length of time they must live with an intestinal stoma. Author Contributions Study conception and design: Jafferji, Hyman Acquisition of data: Jafferji, Hyman Analysis and interpretation of data: Jafferji, Hyman Drafting of manuscript: Jafferji, Hyman Critical revision: Jafferji, Hyman REFERENCES 1. Etzioni DA, Mack TM, Beart RW Jr, Kaiser AM. Diverticulitis in the United States: 1998-2005: changing patterns of disease and treatment. Ann Surg 2009;249:210e217. 2. Masoomi H, Buchberg BS, Magno C, et al. Trends in diverticulitis management in the United States from 2002 to 2007. Arch Surg 2011;146:400e406.
Vol. 218, No. 6, June 2014
3. Rafferty J, Shellito P, Hyman NH, Buie WD. Practice parameters for sigmoid diverticulitis. Dis Colon Rectum 2006;49: 939e944. 4. Khosraviani K, Campbell WJ, Parks TG, Irwin ST. Hartmann procedure revisited. Eur J Surg 2000;166:878e881. 5. Farthmann E, Ru¨ckauer K, Ha¨ring R. Evidence-based surgery: diverticulitisea surgical disease? Langenbeck’s Arch Surg 2000; 385:143e151. 6. Trenti L, Biondo S, Golda T, et al. Generalized peritonitis due to perforated diverticulitis: Hartmann’s procedure or primary anastomosis? Int J Colorectal Dis 2011;26:377e384. 7. Constantinides VA, Heriot A, Remzi F, et al. Operative strategies for diverticular peritonitis: a decision analysis between primary resection and anastomosis versus Hartmann’s procedures. Ann Surg 2007;245:94e103. 8. Zorcolo L, Covotta L, Carlomagno N, Bartolo D. Safety of primary anastomosis in emergency colo-rectal surgery. Colorectal Dis 2003;5:262e269. 9. Breitenstein S, Kraus A, Hahnloser D, et al. Emergency left colon resection for acute perforation. Primary anastomosis or Hartmann’s procedure? A case-matched control study. World J Surg 2007;31:2117e2124. 10. Abbas S. Resection and primary anastomosis in acute complicated diverticulitis, a systematic review of the literature. Int J Colorectal Dis 2007;22:351e357. 11. Constantinides VA, Tekkis PP, Athanasiou T, et al. Primary resection with anastomosis vs. Hartmann’s procedure in nonelective surgery for acute colonic diverticulitis: a systematic review. Dis Colon Rectum 2006;49:966e981. 12. Masoomi H, Stamos MJ, Carmichael JC, et al. Does primary anastomosis with diversion have any advantages over Hartmann’s procedure in acute diverticulitis? Dig Surg 2012;29: 315e320. 13. Wedell J, Banzhaf G, Chaoui R, et al. Surgical management of complicated colonic diverticulitis. Br J Surg 1997;84:380e383. 14. Deans G, Krukowski Z, Irwin S. Malignant obstruction of the left colon. Br J Surg 2005;81:1270e1276. 15. Healey MA, Shackford SR, Osler TM, et al. Complications in surgical patients. Arch Surg 2002;137:611e618. 16. Vermeulen J, Akkersdijk GP, Gosselink MP, et al. Outcome after emergency surgery for acute perforated diverticulitis in 200 cases. Dig Surg 2007;24:361e366. 17. Salem L, Flum DR. Primary anastomosis or Hartmann’s procedure for patients with diverticular peritonitis? A systematic review. Dis Colon Rectum 2004;47:1953e1964. 18. Vermeulen J, Coene PP, Van Hout NM, et al. Restoration of bowel continuity after surgery for acute perforated diverticulitis: should Hartmann’s procedure be considered a one-stage procedure? Colorectal Dis 2009;11:619e624. 19. Aydin HN, Remzi FH, Tekkis PP, Fazio VW. Hartmann’s reversal is associated with high postoperative adverse events. Dis Colon Rectum 2005;48:2117e2126. 20. Garber A, Hyman N, Osler T. Complications of Hartmann takedown in a decade of preferred primary anastomosis. Am J Surg 2014;207:60e64. 21. Roberson I, Eung E, Hughes D, et al. Prospective analysis of stoma related complications. Colorectal Dis 2005;7:279e285. 22. Cataldo P, Hyman N. Ostomy Management. In: Yeo C, ed. Surgery of the Alimentary Tract. 7th edition. Philadelphia: Sanders Elsevier Publishing; 2012:2248e2261. 23. Regenet N, Pessaux P, Hennekinne S, et al. Primary anastomosis after intraoperative colonic lavage vs. Hartmann’s
Welch
24. 25.
26. 27.
28.
29.
30.
31. 32. 33.
Invited Commentary
1161
procedure in generalized peritonitis complicating diverticular disease of the colon. Int J Colorectal Dis 2003;18:503e507. Biondo S, Perea M, Rague´ JM, et al. One-stage procedure in non-elective surgery for diverticular disease complications. Colorectal Dis 2008;3:42e45. Gooszen A, Gooszen H, Veerman W, et al. Operative treatment of acute complications of diverticular disease: primary or secondary anastomosis after sigmoid resection. Eur J Surg 2003;167:35e39. Alanis A, Papanicolaou GK, Tadros RR, Fielding LP. Primary resection and anastomosis for treatment of acute diverticulitis. Dis Colon Rectum 1989;32:933e939. Hoemke M, Treckmann J, Schmitz R, Shah S. Complicated diverticulitis of the sigmoid: a prospective study concerning primary resection with secure primary anastomosis. Dig Surg 1999;16:420e424. Oberkofler CE, Rickenbacher A, Raptis DA, et al. A multicenter randomized clinical trial of primary anastomosis or Hartmann’s procedure for perforated left colonic diverticulitis with purulent or fecal peritonitis. Ann Surg 2012;256:819e827. Hyman N, Osler T, Cataldo P, et al. Anastomotic leaks after bowel resection: What does peer review teach us about the relationship to postoperative mortality? J Am Coll Surg 2009;208: 48e52. Shackford S, Ben-Jacob T, Ratliff J, Hyman N. Is risk adjusted mortality an indicator of quality of care in general surgery? A comparison of risk adjustment to peer review. Ann Surg 2010; 252:452e458. Platz J, Hyman N, Osler T. Tracking intraoperative complications. J Am Coll Surg 2012;215:519e523. Afshar S, Kurer MA, Bretagnol F. Laparoscopic peritoneal lavage for perforated sigmoid diverticulitis. Colorectal Dis 2012;14:135e142. Bretagnol F, Pautrat K, Mor C, et al. Emergency laparoscopic management of perforated sigmoid diverticulitis: a promising alternative to more radical procedures. J Am Coll Surg 2008;206:654e657.
Invited Commentary John P Welch, MD, FACS Bloomfield, CT Complicated sigmoid diverticulitis with free perforation is treated surgically with several different approaches. The Hartmann procedure (HP), popularized in the later 20th century, is gradually being supplanted by primary anastomosis (PA), with or without ileostomy, and by laparoscopic peritoneal lavage and drainage. In this study, the authors compare the outcomes of HP and PA over a 15-year period at 1 center, hypothesizing that colorectal surgeons would be more likely to perform PA than noncolorectal surgeons. This study is representative of contemporary management of a relatively common surgical emergency by a multispecialty group. The comparison groups (noncolorectal
MD,
Neil Hyman,
MD, FACS
The “best” operation in the setting of acute complicated diverticulitis has been debated for decades. Multiple studies, including a recent prospective randomized trial, have reported improved outcomes with primary anastomosis. The aim of this study was to determine whether surgeon or patient-specific factors drives the choice of operative procedure. STUDY DESIGN: Consecutive adult patients with sigmoid diverticulitis, requiring emergent operative treatment for acute complicated diverticulitis, from 1997 to 2012 at an academic medical center, were identified from a prospectively maintained complications database. Patient characteristics, surgeon, choice of operation, and outcomes including postoperative complications and stoma reversal were noted. The use of primary anastomosis and associated outcomes between colorectal and noncolorectal surgeons were compared. RESULTS: There were 151 patients who underwent urgent resection during the study period, and 136 met inclusion criteria. Eighty-two resections (65.1%) were performed by noncolorectal surgeons and 44 by colorectal surgeons (34.9%). Noncolorectal surgeons performed more Hartmann procedures (68.3% vs 40.9%, p ¼ 0.01) despite similar demographics, American Society of Anesthesiologists (ASA) classification, and Hinchey stage. Length of stay, time to stoma reversal, ICU days, and postoperative complications were lower in the colorectal group (43.2% vs 16.7, p ¼ 0.02). CONCLUSIONS: Although patient-specific factors are important, surgeon is a potent predictor of operation performed in the setting of severe acute diverticulitis. A more aggressive approach to primary anastomosis may lower the complication rate after surgical treatment for severe acute diverticulitis. (J Am Coll Surg 2014;218:1156e1162. 2014 by the American College of Surgeons)
BACKGROUND:
More than 300,000 inpatient admissions for diverticulitis occur annually in the US.1 Although most cases of diverticulitis may be treated successfully without surgery, severe disease often requires emergent or urgent surgical management.2 For decades, the Hartmann procedure (HP) has been considered by many to represent the standard surgical approach for complicated diverticulitis.3-6 However, a growing number of studies have suggested that most acute cases may be safely treated by primary anastomosis (PA) with or without a diverting ileostomy.6-11 Despite the abundant literature attesting to the safety of primary anastomosis, HP continues to be the most commonly chosen alternative operation performed Disclosure Information: Nothing to disclose. Presented at the New England Surgical Society 94th Annual Meeting, Hartford, CT, September 2013. Received September 25, 2013; Revised November 6, 2013; Accepted December 9, 2013. From the Department of Surgery, University of Vermont College of Medicine, Burlington, VT. Correspondence address: Neil Hyman MD, FACS, Fletcher 465, University of Vermont College of Medicine, Burlington, VT 05401. email: Neil. [email protected]
ª 2014 by the American College of Surgeons Published by Elsevier Inc.
in the acute setting in the United States.1,12 The HP requires at least 2 major intra-abdominal operations, and many patients never have reversal of their stoma.13,14 Patients with complicated diverticulitis are typically seen in the emergency department and are evaluated by on-call surgeons, who determine the treatment of choice based on clinical and radiologic assessment, as well as their training and experience. Our aim was to assess whether surgeon or disease-specific characteristics primarily drive the type of operation performed in the setting of severe, acute complicated diverticulitis. To achieve this aim, we divided the surgeons who take night call for adult surgical emergencies into 2 categories: colorectal surgeons vs noncolorectal surgeons. We hypothesized that colon and rectal surgeons would be more likely to perform primary anastomosis. We also aimed to determine if any observed variation in surgical practice was associated with differences in key outcomes.
METHODS Study design and subject selection We conducted a retrospective chart review of all adult patients admitted through the Emergency Department
1156
ISSN 1072-7515/14/$36.00 http://dx.doi.org/10.1016/j.jamcollsurg.2013.12.063
Vol. 218, No. 6, June 2014
Jafferji and Hyman
Abbreviations and Acronyms
ASA HP LOS PA PA-DI
¼ ¼ ¼ ¼ ¼
American Society of Anesthesiologists Hartmann procedure length of stay primary anastomosis primary anastomosis with diverting ileostomy
of Fletcher Allen Healthcare, the teaching hospital of the University of Vermont College of Medicine, from January 1, 1997 through December 31, 2012, with a diagnosis of acute sigmoid diverticulitis (ICD-9 Diagnosis Code 562.11), who underwent an emergent or urgent surgical procedure during the index hospitalization. Patients were excluded from further analysis if they were under 18 years of age, had diverticulitis other than the sigmoid colon, had an alternative diagnosis made based on final histopathologic examination, a history of previous colon resection, or instances in which the admitting surgeon differed from the operating surgeon. Patients who did not have clear radiologic and/or intraoperative findings of perforated diverticulitis (defined as Hinchey stages 1 to 4) and those who underwent a radiologic drainage procedure before surgery or laparoscopic lavage were also excluded. Demographic information including age, sex, and body mass index (BMI) were noted. The white blood cell count and heart rate on admission to the emergency department were recorded. American Society of Anesthesiologists (ASA) classification was obtained from preoperative anesthesia records; Hinchey stage and any intraoperative complications were abstracted from the dictated operative report. The admitting/operating surgeon was recorded, and specific surgeon characteristics, including fellowship training and years in practice, were obtained from the institution’s surgical department database. Surgical procedures were classified as a Hartmann procedure (HP), resection with primary anastomosis (PA), or primary anastomosis with diverting ileostomy (PA-DI), based on review of the operative report. Postoperative complications were obtained from the Surgical Activity Tracking System (SATS), a prospectively maintained and validated complication database. Details of this tracking system have been published elsewhere.15 Any admission to the ICU, reoperation, or postoperative drainage procedure by interventional radiology was tracked in the patient’s medical record. The length of stay (LOS) was obtained from the patient discharge summary. Outpatient clinic records were routinely reviewed to capture complications occurring after discharge and to identify stoma reversal and days to reversal as applicable.
Surgery for Complicated Diverticulitis
1157
Statistical analysis Categorical variables such as ASA class, Hinchey class, or operation type were subjected to chi-square analysis and/ or Fisher’s exact test. Continuous data such as age, length of stay, length of operation, white blood cell count, and heart rate were analyzed with a Student’s t-test. Statistical computation was achieved using MS Excel 2007 and Stata/MP 12.0. The study was approved by the Institutional Review Board of the University of Vermont College of Medicine.
RESULTS A total of 151 patients were admitted through the emergency department with a diagnosis of severe acute complicated diverticulitis and underwent a surgical procedure at index hospitalization during the study period. Fifteen patients were excluded, usually based on disease location, subsequent histopathologic review, or the use of preoperative radiologic drainage; 10 patients were excluded because they were treated with laparoscopic washout without resection (Fig. 1). All procedures were performed by 1 of 14 board-certified surgeons who took general surgery night call during the study period; 3 were general surgeons, 4 were fellowship-trained colon and rectal surgeons, and 7 had fellowship training in other specialties (minimally invasive surgery, surgical oncology, or trauma). The surgeons were in practice a median of 11 years (range 1 to 27 years) at the time of colectomy. There were 88 operations (64.7%) performed by general surgeons or noncolorectal fellowship-trained surgeons; 48 operations (35.3%) were performed by a colorectal surgeon. Patient characteristics including Hinchey stage were similar between the 2 groups (Table 1). A total of 10 patients were treated with laparoscopic washout and were excluded from further analysis. Noncolorectal surgeons performed HP in 56 of 82 of the resected cases; colorectal surgeons performed HP in 18 of 44 instances (68.3% vs 40.9%, p ¼ 0.01). Patients in both groups who underwent a PA or PA-DI were significantly younger (p ¼ 0.04) and had a lower Hinchey class (p ¼ 0.01) and a lower ASA status (p ¼ 0.03) than those who underwent HP. The outcomes in the patients cared for by the noncolorectal and colorectal surgeons are displayed in Table 2 and by operation in Table 3. Of the Hinchey stage 4 patients, 16 of 18 noncolorectal patients underwent HP vs 3 of 8 in the colorectal group. In general, there was no difference in duration of surgery, overall LOS, need for reintervention, and the percentage of patients ultimately undergoing stoma reversal. Patients in the colorectal group were less likely to have an ICU admission, had a shorter LOS,
1158
Jafferji and Hyman
J Am Coll Surg
Surgery for Complicated Diverticulitis
Figure 1. Patient population.
and fewer days to stoma reversal. There were 13.7% of patients who required reintervention by interventional radiology or reoperation in the noncolorectal group vs 8.3% in the colorectal group (p ¼ 0.07). There were 2 intraoperative complications among the general group and none in the colorectal group. One was a bladder injury and the other was a ureteral injury. Postoperative complications occurred in 43.2% and 16.7% in the general and colorectal groups, respectively (p ¼ 0.02). There was 1 death in the noncolorectal group. The HP had approximately twice the complication rate of either PA group (Table 3).
Table 1.
DISCUSSION We found that colorectal surgeons were more likely to perform primary anastomosis in the setting of acute complicated diverticulitis than noncolorectal surgeons, despite operating on similar patients. This more aggressive approach was associated with fewer complications, shorter length of stay, elimination of ICU stays, and a shorter time to stoma reversal. However, it was not our primary aim or intention to compare the proficiency of colorectal surgeons to general or other fellowship specialty surgeons, nor to necessarily claim any sort of superiority in care based on training.
Study Patient Characteristics, n ¼ 136
Characteristics
Median age, y Female, % Median body mass index, kg/m2 Hinchey classification, median, % I II III IV
General and other surgeons
Colorectal surgeons
p Value
61 36.4 27.8 3.0 2 22 46 18
56 50.0 27.6 3.0 4 12 24 8
0.32 0.07 0.62 0.96
Vol. 218, No. 6, June 2014
Table 2.
Jafferji and Hyman
General and other surgeons
Colorectal surgeons
p Value
8.5 20.5 13.7 181 61.4 124
7 0 8.3 169 68.2 92
0.86 0.03 0.07 0.63 0.24 0.04
Median length of stay, d ICU admission, % Reintervention, % Median length of operations, min Stoma reversal, % Stoma reversal, d
It may be the case that other groups of surgeons produce equivalent or even superior outcomes in patients with complicated diverticulitis. Certainly, the small number of surgeons overall and in each subcategory would preclude any confidence in making such a sweeping assertion. Rather, we sought to use this comparison to help discern whether surgeon-specific or patient factors are the primary drivers of the operative approach in the setting of acute complicated diverticulitis. Undoubtedly, both the condition of the patient and the experience of the surgeon play some role in surgical decision making. Not surprisingly, patients who were older, had more advanced local sepsis, and a higher ASA classification were more likely to undergo an HP than their younger, more fit counterparts. This is the sort of observation that typically confounds the interpretation of the comparative studies of HP with PA.16,17 But our finding that more than twothirds of patients with complicated disease underwent colorectal anastomosis in the colorectal group vs well less than half of a well matched cohort, suggests that surgeon characteristics are a potent predictor of the operation performed. Although most patients with PA in this series had a protective loop ileostomy (PA-DI), the time to stoma closure was much shorter than for those who underwent
Variable
1159
Outcomes of Surgery, n ¼ 126
Characteristics
Table 3.
Surgery for Complicated Diverticulitis
HP. Hartmann takedown requires a second major laparotomy with a significant risk of inadvertent enterotomy and serious complications.18,19 In a relatively large series from our institution, the overall postoperative morbidity of Hartmann takedown was 29%, with an associated mortality of 2%.20 Further, approximately 90% of patients with PA-DI in previous comparative studies undergo ileostomy closure, and up to one-half of HP patients end up with a permanent stoma.9,18 In this series, performing primary anastomosis did result in somewhat longer operative times, especially when a loop ileostomy was added, but there was no difference in operative times between the colorectal and noncolorectal surgeons overall. The somewhat longer time to perform the anastomosis and construct the loop ileostomy must be weighed against the potential challenges of colostomy creation. Creating a left-sided colostomy in the acute setting, especially in obese patients, can be challenging and painstaking.21 Several patients in the HP group did require reoperation for colostomy complications, generally ischemia and/or retraction. Performing an anastomosis is often technically easier, and there is usually less trouble creating a loop ileostomy than an end descending colostomy.22
Comparison of Operative Groups by Surgeon Training, n ¼ 126
HP
Surgeon training Noncolorectal Colorectal PA PA-DI HP PA PA-DI
Median age, y 63.8 58 57.5 59.7 53.9 Median BMI, kg/m2 26.7 26.5 29.6 29.6 27.5 Median admission heart rate, beats/min 102 95 98 108 96 Median admission white blood cell count, 103/mL 14 9 11 13 10 Median Hinchey class 3.18 2.67 3 2.8 2.3 Median ASA class 3 3 2.8 3.1 2.7 Median length of operation, min 141 186 217 136 177 ICU admission, % 17.5 0 3 0 0 Mean length of stay, d 10.5 10.5 14 13.5 5.5 Complications, n (%) 19 (22) 9 (10.5) 10 (11.7) 4 (9) 3 (5.7) Median stoma, d 170 e 118 109 e
55.7 25.8 109 12 3.2 2.5 194 0 6.7 1 (2) 77
p Value (noncolorectal vs colorectal) HP PA PA-DI
0.29 0.68 0.69 0.87 0.23 0.88 0.12 0.03 0.01 0.06 0.03
0.45 0.78 0.76 0.67 0.44 0.35 0.48 0.79 0.04 0.08 e
0.23 0.39 0.31 0.59 0.76 0.40 0.26 0.22 0.01 0.06 0.008
ASA, American Society of Anesthesiologists; BMI, body mass index; HP, Hartmann procedure; PA, primary anastomosis; PA-DI, primary anastomosis with diverting ileostomy.
1160
Jafferji and Hyman
Surgery for Complicated Diverticulitis
There were more than twice as many postoperative complications in the noncolorectal group overall. This seems to be partially related to choice of operation because HP had the highest complication rate among the 3 operations, but also is related to surgeon-specific factors; there was a clear trend toward higher complications in the noncolorectal group for each individual procedure studied, as demonstrated in Table 3. We were surprised that no patient in the colorectal group required a stay in the ICU. The ICU stays were largely restricted to the noncolorectal patients who underwent an HP. The HP is often relegated to patients with the worst septic complications and/or the highest risk patients, so this finding would be expected.16 Certainly, patients with free diverticular perforation can be hemodynamically unstable, and we would not wish to imply that intensive care stays can be eliminated in patients with complicated diverticulitis. Rather, it appears that most patients do not present in this manner and can be safely managed without an ICU stay. The HP remains the most frequently chosen option in the United States for patients requiring acute surgery for diverticulitis. In a review of almost 100,000 patients from the National Inpatient Sample, more than 57% of patients were treated with a colostomy.12 Despite the abundant evidence that has accumulated over recent years attesting to the safety of primary anastomosis in the acute setting, there has not been a corresponding increase in the use of PA in actual practice.1,12 Multiple case series have reported superior outcomes for patients undergoing PA.6-11,17,23-27 However, as we observed in our series, these analyses are typically plagued by selection bias because healthier patients with less complicated disease are preferentially selected to have PA. A prospective randomized trial of 62 well matched patients comparing HP with PA-DI for the management of perforated left colonic diverticulitis demonstrated similar outcomes after the initial colectomy but revealed fewer serious complications in the PA-DI group (0% vs 20%), less cost, shorter operative times, and shorter length of stay associated with the secondary stoma closure procedure. Further, the stoma reversal rate was 90% in the PADI group vs 57% with HP.28 We did not break down the stoma reversal rate by individual operation in this series. The HP can be a lifesaving operation and likely has an ongoing role in managing selected patients with complicated diverticulitis. For example, patients who are hemodynamically unstable before surgery may present difficulties in the diagnosis of a postoperative anastomotic leak and may not be able to tolerate a second septic insult if a leak occurs. Further, there seems little reason to subject a patient with disabling fecal incontinence or those who are
J Am Coll Surg
institutionalized with serious comorbid conditions to the risk of anastomosis. Indeed, even the colorectal surgeons applied this operation in almost one-third of cases. However, as a rule, there may be little justification for HP in many patients with complicated diverticulitis. There are several limitations to this study. First this was a retrospective study at 1 institution involving a limited number of surgeons. A multicenter study would help to increase generalizability as well as increase the statistical power of the comparison. However, we also believe that using only our patients, whose outcomes are recorded by a rigorous, standardized prospective complication tracking system, also has some very real benefits regarding the completeness, accuracy, and uniformity of identifying adverse events.29-31 In addition, the decision to perform PA is not the only dilemma in the present day management of patients with severe acute diverticulitis. It remains unclear whether or not a loop ileostomy is really required after sigmoid resection with anastomosis,6,23 and we did not address the role of laparoscopic washout in patients requiring urgent surgery for perforated diverticulitis.32,33
CONCLUSIONS In conclusion, we believe that there is an opportunity for improvement in the care of patients requiring acute surgery for complicated diverticulitis by the more liberal use of PA, with or without diverting loop ileostomy. Although the HP likely has an ongoing role in the surgical armentarium and our data do not provide a basis for deciding when it is best applied, there seems little doubt that PA is underused. A more aggressive stance toward PA in the setting of acute complicated diverticulitis may lower complication rates, decrease length of stay, and improve patient satisfaction by reducing the incidence and length of time they must live with an intestinal stoma. Author Contributions Study conception and design: Jafferji, Hyman Acquisition of data: Jafferji, Hyman Analysis and interpretation of data: Jafferji, Hyman Drafting of manuscript: Jafferji, Hyman Critical revision: Jafferji, Hyman REFERENCES 1. Etzioni DA, Mack TM, Beart RW Jr, Kaiser AM. Diverticulitis in the United States: 1998-2005: changing patterns of disease and treatment. Ann Surg 2009;249:210e217. 2. Masoomi H, Buchberg BS, Magno C, et al. Trends in diverticulitis management in the United States from 2002 to 2007. Arch Surg 2011;146:400e406.
Vol. 218, No. 6, June 2014
3. Rafferty J, Shellito P, Hyman NH, Buie WD. Practice parameters for sigmoid diverticulitis. Dis Colon Rectum 2006;49: 939e944. 4. Khosraviani K, Campbell WJ, Parks TG, Irwin ST. Hartmann procedure revisited. Eur J Surg 2000;166:878e881. 5. Farthmann E, Ru¨ckauer K, Ha¨ring R. Evidence-based surgery: diverticulitisea surgical disease? Langenbeck’s Arch Surg 2000; 385:143e151. 6. Trenti L, Biondo S, Golda T, et al. Generalized peritonitis due to perforated diverticulitis: Hartmann’s procedure or primary anastomosis? Int J Colorectal Dis 2011;26:377e384. 7. Constantinides VA, Heriot A, Remzi F, et al. Operative strategies for diverticular peritonitis: a decision analysis between primary resection and anastomosis versus Hartmann’s procedures. Ann Surg 2007;245:94e103. 8. Zorcolo L, Covotta L, Carlomagno N, Bartolo D. Safety of primary anastomosis in emergency colo-rectal surgery. Colorectal Dis 2003;5:262e269. 9. Breitenstein S, Kraus A, Hahnloser D, et al. Emergency left colon resection for acute perforation. Primary anastomosis or Hartmann’s procedure? A case-matched control study. World J Surg 2007;31:2117e2124. 10. Abbas S. Resection and primary anastomosis in acute complicated diverticulitis, a systematic review of the literature. Int J Colorectal Dis 2007;22:351e357. 11. Constantinides VA, Tekkis PP, Athanasiou T, et al. Primary resection with anastomosis vs. Hartmann’s procedure in nonelective surgery for acute colonic diverticulitis: a systematic review. Dis Colon Rectum 2006;49:966e981. 12. Masoomi H, Stamos MJ, Carmichael JC, et al. Does primary anastomosis with diversion have any advantages over Hartmann’s procedure in acute diverticulitis? Dig Surg 2012;29: 315e320. 13. Wedell J, Banzhaf G, Chaoui R, et al. Surgical management of complicated colonic diverticulitis. Br J Surg 1997;84:380e383. 14. Deans G, Krukowski Z, Irwin S. Malignant obstruction of the left colon. Br J Surg 2005;81:1270e1276. 15. Healey MA, Shackford SR, Osler TM, et al. Complications in surgical patients. Arch Surg 2002;137:611e618. 16. Vermeulen J, Akkersdijk GP, Gosselink MP, et al. Outcome after emergency surgery for acute perforated diverticulitis in 200 cases. Dig Surg 2007;24:361e366. 17. Salem L, Flum DR. Primary anastomosis or Hartmann’s procedure for patients with diverticular peritonitis? A systematic review. Dis Colon Rectum 2004;47:1953e1964. 18. Vermeulen J, Coene PP, Van Hout NM, et al. Restoration of bowel continuity after surgery for acute perforated diverticulitis: should Hartmann’s procedure be considered a one-stage procedure? Colorectal Dis 2009;11:619e624. 19. Aydin HN, Remzi FH, Tekkis PP, Fazio VW. Hartmann’s reversal is associated with high postoperative adverse events. Dis Colon Rectum 2005;48:2117e2126. 20. Garber A, Hyman N, Osler T. Complications of Hartmann takedown in a decade of preferred primary anastomosis. Am J Surg 2014;207:60e64. 21. Roberson I, Eung E, Hughes D, et al. Prospective analysis of stoma related complications. Colorectal Dis 2005;7:279e285. 22. Cataldo P, Hyman N. Ostomy Management. In: Yeo C, ed. Surgery of the Alimentary Tract. 7th edition. Philadelphia: Sanders Elsevier Publishing; 2012:2248e2261. 23. Regenet N, Pessaux P, Hennekinne S, et al. Primary anastomosis after intraoperative colonic lavage vs. Hartmann’s
Welch
24. 25.
26. 27.
28.
29.
30.
31. 32. 33.
Invited Commentary
1161
procedure in generalized peritonitis complicating diverticular disease of the colon. Int J Colorectal Dis 2003;18:503e507. Biondo S, Perea M, Rague´ JM, et al. One-stage procedure in non-elective surgery for diverticular disease complications. Colorectal Dis 2008;3:42e45. Gooszen A, Gooszen H, Veerman W, et al. Operative treatment of acute complications of diverticular disease: primary or secondary anastomosis after sigmoid resection. Eur J Surg 2003;167:35e39. Alanis A, Papanicolaou GK, Tadros RR, Fielding LP. Primary resection and anastomosis for treatment of acute diverticulitis. Dis Colon Rectum 1989;32:933e939. Hoemke M, Treckmann J, Schmitz R, Shah S. Complicated diverticulitis of the sigmoid: a prospective study concerning primary resection with secure primary anastomosis. Dig Surg 1999;16:420e424. Oberkofler CE, Rickenbacher A, Raptis DA, et al. A multicenter randomized clinical trial of primary anastomosis or Hartmann’s procedure for perforated left colonic diverticulitis with purulent or fecal peritonitis. Ann Surg 2012;256:819e827. Hyman N, Osler T, Cataldo P, et al. Anastomotic leaks after bowel resection: What does peer review teach us about the relationship to postoperative mortality? J Am Coll Surg 2009;208: 48e52. Shackford S, Ben-Jacob T, Ratliff J, Hyman N. Is risk adjusted mortality an indicator of quality of care in general surgery? A comparison of risk adjustment to peer review. Ann Surg 2010; 252:452e458. Platz J, Hyman N, Osler T. Tracking intraoperative complications. J Am Coll Surg 2012;215:519e523. Afshar S, Kurer MA, Bretagnol F. Laparoscopic peritoneal lavage for perforated sigmoid diverticulitis. Colorectal Dis 2012;14:135e142. Bretagnol F, Pautrat K, Mor C, et al. Emergency laparoscopic management of perforated sigmoid diverticulitis: a promising alternative to more radical procedures. J Am Coll Surg 2008;206:654e657.
Invited Commentary John P Welch, MD, FACS Bloomfield, CT Complicated sigmoid diverticulitis with free perforation is treated surgically with several different approaches. The Hartmann procedure (HP), popularized in the later 20th century, is gradually being supplanted by primary anastomosis (PA), with or without ileostomy, and by laparoscopic peritoneal lavage and drainage. In this study, the authors compare the outcomes of HP and PA over a 15-year period at 1 center, hypothesizing that colorectal surgeons would be more likely to perform PA than noncolorectal surgeons. This study is representative of contemporary management of a relatively common surgical emergency by a multispecialty group. The comparison groups (noncolorectal
